Behavioral model generation for symbolic analysis of analog integrated circuits

A novel behavioral model generation method focused to symbolic analysis of analog integrated circuits is presented. The proposed method is carried out by modeling each MOS transistor (MOST) by their small-signal nullor-based model. It is shown that by using this type of modeling, the signal path (SP) between the output-input nodes can be approached. In this way, MOST that are not included In the SP should be eliminated, allowing to see the separation of the bias MOST and those that are not it. Most important is behavioral model generation of each MOST which provides a reduced set of network equations and it is represented by an admittance matrix where the order is given by the number of nodes of the SP. Therefore, small-signal characteristics of analog circuits can be easily approached. The proposed method is illustrated by computing the fully-symbolic transfer function (TF) of the operational transconductance amplifier (OTA) Miller and compared with HSPICE simulation.